Update release notes for 1.17 #41548
Conversation
|
(rust_highfive has picked a reviewer for you, use r? to override) |
RELEASES.md
Outdated
| names][39761]. [RFC 1682] | ||
| * [`Self` may be included in the `where` clause of `impls`][38864]. [RFC 1647] | ||
| * [When coercing to an unsized type lifetimes must be equal][40319]. That is, | ||
| there is no subtypeing between `T` and `U` when `T: Unsize<U>`. For example, |
| * [`Cell::swap`] | ||
| * [`Cell::replace`] | ||
| * [`Cell::into_inner`] | ||
| * [`Cell::take`] |
There was a problem hiding this comment.
I don't see Cell::swap, Cell::replace, Cell::into_inner, and Cell::take at all in the stable documentation. That's suspicious to me, since it feels like it was then stabilized within less than a cycle; perhaps unintentionally? I'm not sure. I don't know what the procedure around minor library stabilizations is, either. The other stabilized APIs are in the stable documentation, which is why I bring this up.
There was a problem hiding this comment.
There was a problem hiding this comment.
@Mark-Simulacrum afaict that is the case. These were stabilized within a single cycle.
RELEASES.md
Outdated
| existing code generally ignores the type of dead-code unless a type-hint is | ||
| provided; this can cause surprising inference interactions particularly around | ||
| defaulting. The new code uniformly ignores the result type of dead-code. | ||
| * [Tuple-structs with private fields are no longer visible][38932] |
There was a problem hiding this comment.
Based on the linked PR it looks like only the constructors of such structs are private, the types are still public (assuming pub struct …)
RELEASES.md
Outdated
| * [Implement `PartialEq<[A]>` for `VecDeque<A>`][38661] | ||
| * [`HashMap` resizes adaptively][38368] to guard against DOS attacks | ||
| and poor hash functions. | ||
| * [Backtraces have nicer formatting][38165], skipping runtime-internal stack |
There was a problem hiding this comment.
Wasn't the "skipping runtime-internal stack frames" part reverted?
RELEASES.md
Outdated
| * [Optimize vtable loads][39995] | ||
| * [Turn off vectorization for Emscripten targets][39990] | ||
| * [Provide suggestions for unknown macros imported with `use`][39953] | ||
| * [Fix ICES in path resolution][39939] |
alexcrichton
added
beta-accepted
|
r=me when this is good to go |
RELEASES.md
Outdated
| * [rustdoc: fix doctests with non-feature crate attributes][38161] | ||
| * [Make transmuting from fn item types to pointer-sized types a hard | ||
| error][34198] | ||
| * [Higher-ranked lifetimes may not appear solely in associated types][33685]. |
There was a problem hiding this comment.
That's the same as #32330.
aidanhs
added
the
S-waiting-on-author
|
Updated. |
|
{Arc/Rc}::ptr_eq got stabilized with 1.17 too: https://doc.rust-lang.org/beta/std/rc/struct.Rc.html#method.ptr_eq |
|
Thanks @CryZe. Updated. |
|
@brson was this waiting on anything else? Or is this ready to go? |
|
Hm one of the travis jobs is marked as cancelled after 45 min 32 sec? Is that a timeout? I'm retrying it within travis... |
|
Nah this is done. @bors r+ |
|
📌 Commit 8e97693 has been approved by |
|
⌛ Testing commit 8e97693 with merge db1c9eb... |
Update release notes for 1.17 [Rendered](https://github.com/brson/rust/blob/relnotes/RELEASES.md). cc @steveklabnik
|
☀️ Test successful - status-appveyor, status-travis |
alexcrichton
removed
the
beta-nominated
[beta] backports - #41666 - #41578 - #41548 - #41529 r? @alexcrichton
Rendered.
cc @steveklabnik